RCE Cloud

  • Research possible cloud solutions for extending the RCE to the Amazon Web Services (AWS) cloud
  • Architect and document a solution to extend the local RCE
  • Configure the AWS Virtual Private Cloud (VPC) with a Virtual Private Network (VPN)
  • Test the VPC / VPN
  • Develop and configure an Amazon Machine Image (AMI) for Condor worker nodes
  • Develop, test, and deploy software services (using Java, Spring, and the AWS API)
    • Enable custom AWS Cloudwatch metric to track Condor pool utilization
    • Poll and deliver utilization value to AWS Cloudwatch
    • Test whether utilization value exceeds a defined threshold
    • Fire up a Condor worker node using a cool-down period logic
    • Terminate Condor worker nodes as utilization threshold requires
    • Initiate sample Condor test jobs to the pool
  • Develop and integrate deployment modes for development, beta, and production
  • Integrate source code with Git
  • Document source code
  • Design an integrated simulation module for the RCE Cloud node management
  • Implement the integrated simulation module
  • Test the integrated simulation module
  • Execute various simulated workloads
  • Configure local network switch (Juniper) to work with Amazon Web Services (AWS) VPC
  • Work with HUIT personnel on configuring larger higher-throughput switches for network connection with AWS and resource sharing among disparate groups
  • Configure Amazon Machine Image (AMI) to be equivalent to a local RCE server OS image
  • Configure separate development, beta, and production clusters with separate network sharing
  • Organize local storage to isolate volumes visible to AWS servers